A 2025 article explores why climate economic impact estimates vary widely, urging model improvements to better guide climate policy.
Economic projections of the costs of climate change have long ranged from minimal to catastrophic, causing confusion among policymakers and the public. This article explores why estimates of global GDP losses from climate change vary so widely—sometimes projecting as little as a few percentage points of GDP loss at 3°C of warming, and other times suggesting losses as high as 50–60%.
The authors argue that the wide divergence stems from differences in methodologies, assumptions, and missing mechanisms, and they call for urgent reconciliation of these models to produce more robust, usable results.
Key Findings: Three Diverging Methodologies
Structural/Process-Based Models (e.g., DICE, FUND, PAGE)
These models use theoretical economic structures and simulate how climate change affects sectors such as labor, agriculture, and energy demand.
Impact estimates are typically modest. For example, at 3°C warming, structural models predict 1–5% GDP losses.
These models are easier to interpret, but may miss certain damage channels, such as indirect effects on conflict, migration, or biodiversity loss.
Statistical Models (Empirical Approaches)
These models use historical data on temperature variations and economic performance to project how future climate changes will impact economies.
Estimates are much higher than structural models, with some studies predicting up to 55% GDP losses at 3°C.
However, these models are criticized for being highly sensitive to assumptions, prone to over-extrapolation, and lacking a clear mechanistic basis.
Meta-Analyses
These studies combine results from both structural and statistical models to identify broader trends.
While helpful for providing an overview, meta-analyses inherit limitations from both methods and often lack consistency in data and assumptions.
Expert Disagreements
The article highlights a notable divide between economists and physical climate scientists:
Economists tend to produce lower impact estimates (e.g., a few percent of GDP loss).
Climate scientists generally estimate much larger losses, reflecting concern about "fat-tailed" risk distributions and potential tipping points.
Another major point of contention is discount rates—the degree to which future damages are devalued in today's terms. Lower discount rates yield higher present-day costs of climate impacts, as seen in the famous Stern Review, while higher rates tend to downplay long-term impacts.
Missing Elements in Current Models
The authors argue that many models exclude key impact channels, including:
Loss of ecosystem services and biodiversity
Increased conflict and migration
Health impacts beyond temperature-related mortality
Social and political instability
Additionally, most models lack adequate treatment of adaptation and potential feedback loops between systems, leading to either under- or overestimation of impacts.
A Call to Action
The paper outlines steps to advance the field:
Greater transparency on model assumptions and limitations.
Disaggregated data to allow comparison across sectors and regions.
Cross-disciplinary collaboration to incorporate missing mechanisms and improve risk assessments.
Improved uncertainty treatment, including low-probability but high-impact events.
Conclusion: Toward Better-Informed Climate Policy
Until these differences are reconciled, policymakers should treat global economic impact estimates as complementary lines of evidence, rather than definitive forecasts. The article recommends a global research effort—including expert workshops and special reports—to harmonize methodologies and offer decision-makers more actionable insights.
Sign up for our newsletter or connect with us on social media to stay up-to-date with our latest posts and permaculture inspiration.
Explore our inspiring series and posts:
Love the post? Share it with your circle, inspire your people: